Deontic ghe vol with past participle in some varieties of ...

41

Deontic ghe vol with past participle in some varieties

of eastern Veneto

Mariachiara Berizzi – Silvia Rossi

(Università di Padova)

1. Introduction1

Modality is an area of grammar that the languages of the world very often express by means

of periphrastic constructions. Indeed, it is often the case that the same modal value, for

instance, epistemic or deontic, is conveyed in the same language (or in a group of closely

related languages) by a number of strategies sometimes only minimally different. A good

example of this is the use of the functional verb want followed by a past participle to express

deontic necessity, a phenomenon well attested in many Italo-Romance varieties (cf. Rohlfs

1969; Ledgeway 2000).2

In this contribution we focus on some varieties of eastern Veneto that show deontic

vo(l)er (“want”) plus past participle. In these varieties, however, vo(l)er is obligatorily

accompanied by the oblique clitic ghe, as exemplified in (1):

(1) Sta roba a *(ghe) vol fata

this thing Cl.sbj.3.f.sg Cl wants done.f.sg

“This thing needs/ has to be done”

Henceforth, we will refer to this construction as ghe vol + p.p.

1 We would like to thank our informants (Girolamo Antonini, Giorgia Buoso, Federico Mazzonetto, Elvi

Pizzinato, Mauro Zanchetta, Elena Zanella) and Paola Benincà, Davide Bertocci and Guglielmo Cinque for their

suggestions. Many thanks also to Luca Melchior for helping us with the Friulian data. Parts of this contribution

have been presented at the XVI Giornata di Dialettologia, Padua 18 June 2010, at CIDSM 5, Berlin 2-3 July

2010 and at the X Incontro di Dialettologia Italiana, Bristol, 22-24 September 2010. We thank the audiences of

these conferences for their helpful comments. Even though this contribution is the result of the constant

collaboration of the two authors, Mariachiara Berizzi is responsible for sections 2, 3 and 5 and Silvia Rossi for

sections 1 and 4. 2 The phenomenon is not only restricted to Italo-Romance varieties but is present, for instance, in Scottish

English and Midlands American English, see Remberger (2006: 254 and sources quoted therein):

(i) The car wants washed “The car needs washing”

Berizzi - Rossi

42

The phenomenon has been investigated in two varieties in the administrative area of Venice,

namely Fossalta di Piave and La Salute di Livenza (S. Stino di Livenza) and two varieties in

the province of Treviso, Soffratta (Mareno di Piave) and Bocca di Strada (Mareno di Piave).

The data have been collected by interviewing informants on the basis of a questionnaire with

two tasks: in the first part they were asked to give grammaticality judgments on the

acceptability of a set of dialectal sentences, in the second, they were asked to translate a

number of Italian sentences into their own varieties.3

From the observation of the data, however, it emerged quite clearly that in the variety

of Fossalta di Piave (henceforth Fossaltino) the phenomenon shows an interesting degree of

morpho-syntactic variation. We will therefore concentrate on this variety, while the micro-

variation across the abovementioned dialects will be adduced as evidence of the peculiar

behaviour of deontic ghe vol in Fossaltino.

As we will show in detail, Fossaltino displays different types of constructions that can

follow deontic ghe vol. These syntactic possibilities are not found in the neighbouring

varieties but are attested with deontic want in Friulian. A closer comparison with the Friulian

equivalent constructions turns out to be very insightful since it shows that we are dealing with

basically two different types of deontic ghe vol, each to be distinguished by a number of

morpho-syntactic factors such as the nature of the subject, the presence or absence of past

participial agreement and the distributional properties of the clitic ghe. The aim of the present

paper is to provide the analysis of the syntax of these constructions.

The paper is organized as follows: in the next section we give a brief overview of the

extant literature on the phenomenon through the description of the most important analyses

put forward in the past twenty years. In Section 3 we describe the different constructions

involving deontic ghe vol in Fossaltino, hinting, when possible and/or useful, at the

microvariation within the group of dialects we have sampled. In this section we also show that

the different behaviour that Fossaltino displays with respect to the neighbouring varieties is

actually paralleled by Friulian. On the basis of this comparison, we propose a unitary

syntactic analysis of these constructions in Section 4. In this section we also consider the

distribution of the clitic ghe in the ghe vol + p.p. of Fossaltino, in the attempt of understanding

3 It should be noted that ghe vol + p.p. is an optional phenomenon since the informants always have alternative

constructions to express deontic necessity:

(i) - bisogna + infinitive / that-clause , “it is necessary to, that“

- aver da + infinitive “to have to”

- the modal dovere

- ndar “go” + past particple, “something has to be done”

Deontic ghe vol with past participle in some varieties of eastern Veneto

43

whether it should be considered a locative subject along the lines of Tortora (1997). Section 5

presents some conclusions.

2. Deontic Want with past participle in Italo-Romance

As already mentioned in the previous sections, deontic constructions with want and the past

participle of a lexical verb are attested throughout the Italo-Romane area:

(2) a. Basso Polesano, PD (Benincà & Poletto 1997: 102)

El vole magnà

Cl.sbj.3.m.sg wants eaten

“This thing has to be eaten”

b. Friulian (Salvioni 1912, in Ledgeway 2000: 244)

La çhosse la ul fate

The thing Cl.sbj.3.f.sg wants done.f.sg

“This thing has to be done/it is necessary to do this thing”

c. Sardinian (Remberger 2006: 250)

Deu bollu agiudau po fai is iscalas

I want.1.sg helped.m.sg to do the stairs

“I need to be helped up the stairs”

d. Salentino, Maglie, LE (Salvioni 1912, in Ledgeway 2000: 244)

Lu pisce ulia mangiatu stammane

The fish wanted eaten.m.sg this-morning

“The fish had to be eaten this morning”

Rohlfs (1969: §738) reports some other cases of this from a number of Southern dialects.

In the linguistic literature, the phenomenon has received a fair amount of consideration

and several analyses have been proposed. In what follows, we briefly present the most

important of them as they turn out to be important for our analysis of deontic ghe vol in

Fossaltino.

Berizzi - Rossi

44

2.1 Benincà & Poletto (1994, 1997) on deontic want plus p.p. in Basso Polesano

In their studies on the deontic verbs of necessity in Italian and in some dialectal varieties of

Veneto, Benincà & Poletto (1994, 1997) take into consideration the deontic use of want with

a past participle in Basso Polesano, a variety of southern Venetan spoken in Polesine. In this

variety the phenomenon is subject to the following morpho-syntactic restrictions:

(i) It is possible only with simple tenses, cf. (3a) vs. (3b)

(3) a. El vole/voeva/ voria/ vorà magnà

Cl.sbj.3.m.sg wants/want.impf/want.pres.cond/want-fut eaten

“It has/had/would have/will have to be eaten”

b. *El ga volesto magnà

Cl.sbj.3.m.sg has wanted eaten

(ii) It is not possible with non-finite tenses, cf. (4):

(4) *El podaria voler magnà/ *Volendo magnà, ...

Cl.sbj.3.m.sg could want eaten/ Wanting eaten

(iii) It appears generally only at 3rd persons, cf. (5):

(5) *Mi voio petenà/ *Ti te voi petenà /*A volemo petenà /*A volì petenà

I want combed/you Cl.sbj.2.sg want combed/Cl.sgg.1/2.pl want combed

Moreover, they show that, in Italian and in the Venetan dialects, there are at least two types of

verbs to express deontic necessity:

Verbs like It. bisogna “to be necessary” and Ven. toca “touch” that do not have a

subject (neither clitic nor DP) and are followed by an infinitive or by a that-clause;

Verbs like It. andare “go” and Ven vo(l)er “want” that admit only a third person

subject and are both followed by a past participle.

The authors propose that both these types of deontic verbs are functional verbs base generated

in the head of a projection dedicated to the deontic modality of necessity in the functional area

of the clause (IP). The difference between the two types of verbs relies in the fact that the first


45

group of verbs has an inert VP, i.e., no thematic grid, while the second group comprises verbs

that are “parasitic” of the VP of the past participle that follows them.

2.2 Ledgeway (2000) on the want-passives of the Southern varieties

In the Southern Italian dialects Ledgeway (2000) identifies a number of different meanings

for want plus a past participle form which he defines WANT-passives. More specifically, he

observes that want shows up in the following three structures, here exemplified with data

from Cosentino (Ledgeway 2000: 236-37):

I. ECM, in which volere assigns accusative case to the object of the lexical verb:

(6) Mariu vo mannata chira littera

Mario wants sent.f.sg that letter.f.s

“Mario wants that letter (to be) sent!”

II. OC, subject control in which the subject of volere is originally an argument of the

lexical verb:

(7) Mario vo mannatu chira littera

Mario wants sent.m.sg that letter.f.sg

“Mario wants to be sent that letter!”

III. SR, subject raising in which the object of the lexical verb is promoted to subject

position:

(8) Vo mannata chira littera

Wants sent.f.sg that letter.f.sg

“That letter must be sent/needs sending”

Ledgeway (2000) also observes that in the ECM (6) and OC (7) examples, volere expresses

volitional modality and as such it imposes thematic restrictions on the subject which must be

[+animated]. On the other hand, in the SR example in (8), volere expresses deontic modality

and thus it does not assign any thematic roles, while its subject is the object of the participle.

Berizzi - Rossi

46

As for the ECM all the dialects considered show the same pattern. On the other hand,

OC and SR show up differently in the Southern varieties and, on the basis of their

distributional properties, Ledgeway distinguishes two groups of Southern varieties. The first

group is that of Northern Calabria and Salento in which the passive auxiliary essere “to be” is

never present and the agreement on the past participle is controlled by the argument that is

raised to subject position, cf. (7) and (8). The second group is that of the other Southern

dialects in which the passive auxiliary essere is always present and the agreement on the

participle either appears at the default masculine singular form or is controlled by its object,

cf. examples in (9) and (10) from Neapolitan (Ledgeway 2000: 236) that correspond

respectively to (7) and (8):

(9) Mario vô esse mannata chella lettera

Mario wants to be sent.f.sg. that letter.f.sg.

“Mario wants to be sent that letter”

(10) Vô esse mannata chella lettera

wants to be sent.f.sg. that letter.f.sg.

“That letter must be sent/needs sending”

2.3 Remberger (2006) on want as a deontic passive auxiliary

Basing upon the observations and the classification made in Ledgeway (2000), Remberger

(2006) restricts the term WANT-passives to those constructions that are characterised by: (a)

the promotion of the internal argument to subject position and the agreement of want with it;

(b) the absence of an overt passive auxiliary; (c) the deontic, and not volitional, value of want.

Moreover, another crucial characteristic of WANT-passives is the absence of an explicit

external argument with the possibility to reactivate it through a PP, as can be seen in the

following example:

(11) Sardinian (Jones 1993, in Remberger 2006:259)

Sa makkina keret accontzada dae unu meccánicu

“The car has to be fixed by a mechanic”


47

Remberger (2006) analyses want as a true passive auxiliary with a deontic feature base

generated in the head of the projection for tense. In other words, the author considers true

WANT-passives only the SR examples of Ledgeway (2000), cf. (8).

3. Deontic ghe vol in Fossaltino

As shown in the previous section, the construction involving volere and a past participle is

attested in many different varieties of Italy, all displaying a cluster of common properties.

However, it also shows a fair degree of morpho-syntactic variation. In this section we will

present the most relevant characteristics of ghe vol + p.p. in Fossaltino, keeping in mind all

the criteria applied by the different analyses presented in section 2. In this description we will

also point out the main differences of Fossaltino ghe vol w.r.t. the characteristics displayed by

the same phenomenon in the other varieties investigated. As we have already seen, a crucial

feature here is the obligatoriness of the oblique clitic ghe, cf. (1). The subject is expressed by

a clitic subject and, when present, the DP subject co-indexed with the clitic is obligatorily left

or right dislocated, as exemplified in (12):

(12) (Sto bicer,) el ghe vol lavà (, sto bicer)

(this glass,) Cl.sbj.3.m.sg Cl wants washed, (this glass)

“This glass needs washing”

The agreement on the participle is controlled by the clitic subject while the external argument

can be reactivated through a PP, as can be observed in (13):

(13) Sta camisa a ghe vol lavada da to mare

This shirt Cl.sbj.3.sg Cl wants washed.f.sg by your mother

“This shirt has to be washed by your mother”

The characteristics we have outlined so far for the ghe vol + p.p. are evidence for a process of

passivisation (cf. Remberger 2006). This is further confirmed by the fact that the construction

is grammatical only with transitive verbs, as can be seen by the ungrammaticality of (14):

Berizzi - Rossi

48

(14) a. *Mario el ghe vol dormio / teefonà

M. Cl.sbj.3.m.sg Cl wants slept.m.sg / telephoned

b. *Mario el ghe vol partio / ndat

M. Cl.sbj.3.m.sg Cl wants left.m.sg / gone

However, Fossaltino presents another interesting distributional property: ghe vol + p.p. can

also appear with no referential subject at all (neither clitics nor DPs). In this case, the

construction is grammatical not only with transitives (15), but also with unaccusatives (16)

and unergatives (17), (cf. the ungrammaticality of (14)):

(15) Ghe vol netà (la toea) in cusina

Cl wants cleaned (the table) in kitchen

“It is necessary to clean (the table) in the kichen”

(16) Ghe vol ndat de persona

Cl wants gone by person

“It is necessary to go (there) in person”

(17) Ghe vol dormio almanco do ore prima de partir

Cl wants slept at least two hourse before of-leave

“It is necessary to sleep at least two hours before leaving”

According to our informants‟ judgments, the examples in (16) and (17) express a deontic

necessity of a more impersonal nature. This particular interpretation will be dealt in more

details in the following section, in which we will try to account for it in syntactic terms.

As for the morpho-syntactic characteristics observed by Benincà & Poletto (1994,

1997) for Basso Polesano, ghe vol + p.p. in Fossaltino is not subject to the same restrictions.

In particular, it is possible both in the compound (18) and non-finite tenses (19) and it is

marginally possible at the first and second persons plural (20):

(18) Sta roba a g-a voest fata4

This thing Cl.sbj.3.f.sg Cl-has wanted done.f.sg

4 Fossaltino ga is the combination of the auxiliary a “has” and the dative clitic ghe.


49

“It has been necessary to do this thing”

(19) a. Voendoghe firmada sta carta, ...

Wanting-Cl signed this document, …

“Being necessary to sing this document, …”

b. Sta carta qua, a podaria voerghe firmada come no

this document here, Cl.sbj.3.f.sg could want-Cl signed.f.sg as no

“It could be necessary to sign this document or not”

(20) ?Noaltri / voaltri ghe voen / voè petenai

we / you Cl want.1.pl / want.2.pl combed.m.pl

“We/you need to be combed”

The examples in (18) - (20) show that Fossaltino, when compared to the other varieties here

considered, displays a certain degree of variation since the latter are subject to the same

restrictions observed by Benincà & Poletto (1994, 1997) for Basso Polesano. Moreover, in

these same varieties, the construction is grammatical only with transitive verbs and with the

overt expression of the clitic subject, as can be seen in (21):

(21) *Ghe vol magnà / partio / teefonà

Cl wants eaten / left / phoned

Again, the comparison with the other varieties shows that Fossaltino presents not only less

morpho-syntactic restrictions but also a syntactic possibility unknown to the neighbouring

dialects, as illustrated by the contrast between (17) - (19) on one hand and (23) on the other.

However, Fossaltino shows some interesting similarities with Friulian. Friulian, like

all the other varieties here investigated, exhibits a deontic construction with want and a past

participle. Similarly to the Southern dialects a dative clitic is never present, cf. (3b) e (22):

(22) I fruts a vuelin tignûts pe man (Nazzi 2003; s. v. volè)

The kids Cl.sbj.3.m.pl want.3.pl taken-m.pl by hand

“Children have to be held by hand”

Berizzi - Rossi

50

Moreover, Friulian admits the construction with all types of verbs, cf. (23) in which there is

no referential subject clitic but an expletive:

(23) a. Al vûl mangiât

Cl.sbj.3.m.sg / Cl.expl wants eaten

“It is necessary to eat / It has to be eaten”

b. Al vûl tornât

Cl.expl wants returned

“It is necessary to go back”

c. Al vûl durmît

Cl.expl wants slept

“It is necessary to sleep”

The close similarity between the data in (18) - (20) of Fossaltino and the Friulian examples in

(23) suggests that Friulian and Fossaltino have two different types of ghe vol + p.p.:5

i. Deontic ghe vol followed by a passive costruction, cf. (1) and (9). This type of

construction is grammatical only with transitive verbs and the past participle

agrees with the clitic subject. As to the DP subject, if it is expressed, it is always

right or left dislocated.

ii. Deontic ghe vol followed by a participial complement, cf. (12) - (14). In this case

the construction is grammatical with all types of verbs and the past participle

appears in the default masculine singular form. The subject is never expressed by a

5 D. Bertocci (p.c.) points out to us that the deontic periphrastic passive construction of Latin has distributional

properties that closely resemble those of Fossaltino and Friulian want + p.p. In particular, (examples from A.

Ernout & F. Thomas, 1954, Syntaxe latine, Paris, Klincksieck):

with transitive verbs the object is inflected for nominative and the gerund agrees with it while the external

argument can appear in dative:

(i) mihi colenda est virtus

me.dat practicing.f.sg.nom is virtue.f.sg.nom (“I need to practice virtue”)

If no subject is expressed the gerund appears in the default neuter singular form and the construction is

possible with all types of verbs (and, most importantly, even if rarely, an object can appear in the

accusative):

(ii) nunc est bibendum

Now is drinking.neut.sg.nom (“Now it is time to drink”)

(iii) veniendum est

coming-neut.sg.nom is (“It is necessary to go”)

(iv) viam quam nobis quoque ingrediendum sit

way.f.sg.acc which.f.sg.acc us.dat also enter.neut.sg.nom be (“the road we also need to enter”)


51

referential clitic subject but either it is expressed by an expletive (Friulian) or it is

absent (Fossaltino).

The following examples from Fossaltino further illustrate this distinction:

(24) deontic ghe vol + passivisation

(Sta torta,) a ghe vol magnada (, sta torta)

This cake-fem Cl.sbj.3.f.sg Cl wants eaten.f.sg (this cake)

“This cake has to be eaten”

(25) deontic ghe vol + participial complement

a. Ghe vol magnà sta torta

Cl wants eaten this cake

“It is necessary to eat this cake”

b. *A ghe vol magnà sta torta6

Cl.sbj.3.f.sg Cl wants eaten this cake

c. STA TORTA ghe vol magnà

this cake CL wants eaten

“It is this cake that we have to eat”

The absence of the agreement on the participial in (25a) indicates that sta torta is the direct

object of the lexical verb, while in (24) the same DP originates as the internal argument of the

lexical verb but is promoted to subject position by a process of passivisation, of which the

participial agreement is a manifestation. The example in (25b) shows that when a referential

subject clitic is present the ghe vol + participial complement construction is not possible.

(25c) shows moreover that, in this context, the DP sta torta can be moved to sentence initial

position but it is only interpretable as a contrastive Focus,

6 The fact that the sentence in (i) is marginally acceptable does not necessarily follow from the ungrammaticality

of (25b):

(i) Ghe vol magnada sta torta

Cl wants eaten3.f.sg this cake

More precisely, (25b) shows that if the referential subject clitic is present, there should also be participial

agreement, while the opposite is not necessarily true, i.e., the presence of the subject clitic depends on the more

or less obligatory realisation of a subject clitic in that variety.

Berizzi - Rossi

52

Finally, Fossaltino deontic ghe vol can be followed by a that-clause, a characteristic

that this variety shares not only with some of the other varieties here investigated7 but also

with Friulian, cf. (27):

(26) G-a voest che te ciamesse zinque volte parchè te te giresse

Cl-has wanted that Cl.obj.2.sg call-1sg.pst five times because you yourself turn

“It has been necessary that I called you five times before you turned round”

(27) Se il Friûl, (…), nol vûl deventâ un impuartadôr di sanc, (…), al vûl che

if the F. not-Cl.sbj.3.m.sg wants to-become an importer of blood, Cl.expl wants that

ogni donatôr al fâsi al mancul dôs donazions ad an.

each donor Cl.sbj.3.m.sg make at least two donations per-year

“If the Friuli region (…) does not want to become an importer of blood, it is necessary

that each donator should give blood at least twice a year”

(http://www.friul.net/archivi_gnovis.php?r_mese=04&r_anno=2010&i=2)

4. For an analysis of deontic ghe vol constructions in Fossaltino

As we have seen in the previous section, Fossaltino displays a wide range of syntactic

possibilities for deontic ghe vol, i.e., it can be followed by (i) a passive construction, (ii) a

participial complement, and (iii) a that-clause. In what follows, we will provide a syntactic

analysis that aims to account for such a distribution in a unitary way.

Following the proposal of Cinque (1999), the different modal values (evidential,

epistemic etc.) are encoded in distinct functional projections ordered in a universal hierarchy,

here reported in a simplified way for what concerns modality:

(28) [allegedly ModEvidential [probably ModEpistemic [once T(past) [then T(future)

[perhaps MoodIrrealis [necessarily ModNecessity [possibly ModPossibility

[intentionally ModVolitional [ModObligation [ModAbility/Permission ...

7 This possibility is attested also in the Trevisan varieties investigated, while it is absent in the eastmost variety,

S. Stino di Livenza (VE).

http://www.friul.net/archivi_gnovis.php?r_mese=04&r_anno=2010&i=2


53

Moreover, Patruno (2005) has claimed that the different modal values attested for want in the

Italo-Romance varieties depend on the functional nature of want, i.e., it is always base

generated in one of the modal projections of (28). The semantic and morpho-syntactic

differences depend on which projection want has been base-generated in.

In the constructions we have examined so far, ghe vol always expresses deontic

modality and thus it is to be analysed as a functional verb base generated in ModNecessity.

However, if want is always base generated in the same projection, it is still to be explained

how the different syntactic possibilities are to be accounted for.

It is worth noticing that the various constructions attested with deontic ghe vol can be

grouped on the basis of the nature of their grammatical subject: deontic ghe vol + passive

construction (i) admits only referential (3rd

person) subjects, while deontic ghe vol +

participial complement (ii) and deontic ghe vol + that-clause (iii) do not have referential

subjects at all (no subject in Fossaltino and an expletive in Friulian). Moreover, the same

distinction emerges also when testing the different distributional properties of the clitic ghe in

each construction.

Benincà & Tortora (2009) discuss the properties of the clitic ghe of the Venetan

vo(l)erghe “want-ghe” “it is necessary”. They show that this verb is ambiguous between two

readings: the first is purely deontic (“it is necessary”), while the second is deontic and

benefactive, i.e., it means “to be necessary for someone”, as can be seen in the two English

translations given for (29), respectively (29a) and (29b):

(29) Ghe vole do euro

Cl wants two euros/to-him want two euros

a. “Two euros are necessary”

b. “Two euros are necessary for him/her/them = s/he/they need(s) two euros”

The benefactive reading becomes more evident when we try to introduce a first person

benefactive clitic me. In this case, ghe cannot co-occur with the benefactive clitic, as (30)

shows:

(30) Me (*ghe) vole do euro

“I need two euros”

Berizzi - Rossi

54

Thus, they claim that ghe in vo(l)erghe can be either a deontic clitic, ghedeon, or a dative

(benefactive) clitic, ghedat, both having the capability of licensing the head of the projection

for deontic modality (ModNecessity in Cinque‟s 1999 terms), occupied by vole “wants”.

We can try and see now whether ghe in all the constructions we have described so far

is either a ghedeon or a ghedat by observing what happens when we introduce a first person

dative (benefactive) clitic me. First, in all the three deontic ghe vol constructions the

benefactive me and the clitic ghe cannot co-occur, as Benincà & Tortora (2009) has shown

for vo(l)erghe:

(31) a. ghe vol + passive construction

*A me ghe vol lavada drioman, sta camisa

Cl.sbj.3.f.sg Cl.dat.1.sg Cl wants washed.f.sg immediately this shirt

b. ghe vol + participial complement

*Me ghe vol telefonà al dotor

Cl.dat.1.sg Cl wants phoned to-the doctor

c. ghe vol + that-clause

*Me ghe vol che vae casa drioman

Cl,.dat.1.sg Cl wants that go.1.sg home immediately

However, there is an interesting asymmetry: while in the case of ghe vol + passive

construction, (i), the clitic ghe is in complementary distribution with the benefactive me, cf.

(32), in the case of both ghe vol + participial complement (ii) and ghe vol + that-clause (iii),

the only admitted clitic is ghe, cf. (33):

(32) A me vol lavada drioman, sta camisa qua

Cl.sbj.3.f.sg Cl.dat.1.sg wants washed-f-sg immediately, this shirt here

“I need this shirt washed immediately”

(33) a. *Me vol partio / dormio / ndat

Cl.dat.1.sg wants left / slept / gone

b. *Me vol che vae casa drioman

Cl.dat.1.sg wants that go.1sg home immediately


55

The contrast between (32) and (33) tells us that only in the case of ghe vol + passive

construction can ghe have a benefactive reading (it can be both ghedat and ghedeon) while with

ghe vol + participial clause/that-clause ghe is only ghedeon in Benincà & Tortora‟s (2009)

terms, i.e., a clitic that licenses the head of ModNecessity. However, the exact nature of ghedeon

in all the deontic ghe vol constructions is still to be syntactically identified.

Following very recent work by Parry (2010), ghe could signal the presence of a non

canonical subject, more specifically, of a null locative subject, LOC or pro-loc (cf. Anderson

1980; Freeze 1992; Tortora 1997). Yet, Fossaltino (34) does not show a clitic ghe in the

contexts that Tortora (1997) individuated for Borgomanerese ngh, i.e., existentials and

unaccusative GOAL-entailing verbs, (35), in which ngh is analysed as a locative subject:8

(34) a. E dei bocie in mezo a strada (Fossaltino)

Is/are of-children in middle the road

“There are some kids on the road”

b. E rivà dei bocia/putei

is/are arrived some children

“Some children have arrived”

(35) a. Ngh è-gghi tre mataj int la stônza (Borgomanerese)

Cl.loc.sbj is-Cl.loc three boy in the room

“There are three boys in the room”

b. Ngh è rivà - gghi na fjola

Cl.loc.sbj is arrived-Cl.loc a girl

“A girl has arrived”

(34) shows us that Fossaltino does not have a locative subject ghe in the most prototypical

presentational contexts, thus we suggest that ghe in deontic ghe vol should be given a

different account.

A further comparison with Friulian turns out to be very insightful. As we have seen is

the previous section, Friulian presents a deontic want + past participle construction with all

types of verbs when the subject clitic is the expletive al (recall 23a-c, here repeated in 36).

8 For a typological perspective see Freeze (1992).

Berizzi - Rossi

56

(36) a. Al vûl mangiât

Cl.sbj.3.m.sg / Cl.expl wants eaten

“It is necessary to eat / It has to be eaten”

b. Al vûl tornât

Cl.expl wants returned

“It is necessary to go back”

c. Al vûl durmît

Cl.expl wants slept

“It is necessary to sleep”

The corresponding Fossaltino examples with deontic ghe vol have no subject at all, cf. (18)-

(20), here repeated in (37):

(37) a. Ghe vol netà (la toea) in cusina

Cl wants cleaned (the table) in kitchen

“It is necessary to clean (the table) in the kitchen”

b. Ghe vol ndat de persona

Cl wants gone by person

“It is necessary to go (there) in person”

c. Ghe vol dormio almanco do ore prima de partir

Cl wants slept at least two hourse before of-leave

“It is necessary to sleep at least two hours before leaving”

Notice furthermore that Fossaltino does not have an overt subject in other contexts where

Friulian presents the expletive al, like weather verbs, raising verbs, unaccusatives and

existentials cf. (38) with (39):

(38) a. Al plòuf (Cordenons, PN; ASIt 2.1)

Cl.expl rains (“It rains”)

b. Al somea c' al aipi sigat qualchidun (Cordenons, PN; ASIt 3.19)

Cl.expl seems that Cl.espl has cried someone (“It seems that somebody cired”)

c. Al riva un fivuòl (Cordenons, PN; ASIt 2.18)

Cl.espl arrives a boy (“A boy is appraoching”)

d. Al è un fivuòl (Cordenons, PN; ASIt 2.23)


57

Cl.espl is a boy (“There‟s a boy”)

(39) a. Piove (Fossaltino)

Rains.

b. Par che Piero rive doman

seems that P. arrives tomorrow

c. Riva un bocia

arrives a boy

d. É un bocia

is a boy (“C'è un bambino”)

This similarity leads us to suggest that in the case of deontic ghe vol + participial

complement, Fossaltino displays a null expletive so that the examples in (39) can be analysed

as (40):

(40) a. EXP ghe vol netà (a toea) in cusina

b. EXP ghe vol ndat

c. EXP ghe vol dormio

This analysis can be further extended to the cases of deontic ghe vol + that-clause given the

similarity between (26) and (27) above, so that (26) can be analysed as (41): 9

(41) EXP ghe vol che vae casa drioman

We further suggest that it is the presence of this null expletive that gives to the two

constructions the impersonal reading, while in the case of deontic ghe vol + passive

construction the subject position is filled by a referential clitic in both Fossaltino and Friulian.

9 In these cases ghe vol is like It. bisogna with which some Italian dialects show an expletive clitic:

(i) U bzogna poorti (Altare, Savona; ASIt 2.3)

(ii) Al gh‟è de partir (Semogo, Valdidentro, Sondrio; ASIt 2.3)

“It is necessary to leave”

Both these varieties have the expletives u and al for weather verbs, raising predicates, unaccusative verbs and

existentials. For other cases of the deontic verb toccare meaning bisogna with an expletive subject see Berizzi

(2011).

Berizzi - Rossi

58

As for the nature of the clitic ghe we are still in no position to give a precise definition

of it, yet we tend to consider it a deictic/locative clitic. We leave this most interesting topic

for future research.

Let us now turn to the three types of constructions that follow deontic ghe vol.

In the case of ghe vol + passive construction, Fossaltino — as the Group 1 dialects of

Ledgeway (2000) — does not show the auxiliary passive essere, cf. (42).

(42) Sto bicer, el ghe vol (*essar) lavà

Nonetheless, we assume that it is still present in a silent form given the comparison with the

facts of Group 2 (cf. Ledgeway 2000), cf. (43):

(43) Sto bicer el ghe vol ESSAR lavà

Contra Remberger (2006), we assume here that want is not a passive auxiliary with a deontic

feature: we propose that want is simply a deontic modal verb while the passive meaning is

maintained by a passive auxiliary, even if not overtly lexicalized.

In the case of ghe vol + participial complement we suggest that there should always be

a silent auxiliary but not a passive one: we claim that it is the auxiliary selected by the lexical

verb and its combination with the past participle conveys rather an aspectual value:

(46) a. EXP ghe vol AVER netà (la toea) in cusina

b. EXP ghe vol ESSAR ndat de persona

c. EXP ghe vol AVER dormio almanco …

In other words, we suggest that the construction with the silent auxiliary and the past

participle in (46) conveys the meaning that it is necessary that the event espressed by the

lexical verb (plus its arguments) has to be completed indipendently from whom actually

carries out the event (cf. the arbitrary interpretation given to the subjects of the lexical verbs).

Finally, in the case of ghe vol + that-clause we follow Cinque (2004)‟s and Patruno

(2005)‟s proposal for volitional want and we propose that also in this case ghe vol remains a


59

functional verb and it is then followed by a silent verb that takes a whole CP as its

complement:10

(47) EXP ghe vol NULL VERB che …

As a final note we briefly consider the Friulian cases deontic want plus past participle. We

have argued for a null expletive subject in Fossaltino ghe vol + participial complement and

ghe vol + that-clause on the basis of the equivalent Friulian examples presenting the

lexicalized expletive al. This proposal has been further supported by the fact that Fossaltino

systematically presents no subject in the other contexts in which Friulian presents the

expletive al (existentials, raising verbs, presentational structures with unaccusatives and also

weather verbs).

If we consider the clitic ghe, we have seen that Fossaltino does not present ghe in existentials.

This clitic, however, is not completely absent in the clitic system of Fossaltino: it is present in

the constructions we have examined and as referential locative clitic as in (47):

(47) Ghe vae mi, in posta

Cl.loc go.1.sg. me in post office

“I‟ll go to the post office”

In most Friulian varieties, the clitic ghe is absent even in these contexts:

(48) Tu devis vignì angia tu Moimacco, UD; ASIt 1.17

you must come also you

“You have to come too”

We propose that the Friulian examples in (2b) and (36c) here repeated in (49a) and (50a)

should be analysed as in (49b) and (50b), following a suggestion by P. Benincà (p.c.):

(49) a. Al vûl durmît

10

Cruschina (2010) analyses some modal adverbs of Sicilian that derive from a process of fusion and

grammaticalisation of a verb and the complementiser ca (dicica lit.” say-that”, “presumably”, parica lit.” seems-

that”, “allegedly”, penzica lit.”thinks-that”,” probably”). Ghe vol che ... cannot be analysed as a modal adverb

since the verb want can be inflected and can be separated from the complementiser by low adverbs, differently

from what happens with the modal adverbs of Sicilian.

Berizzi - Rossi

60

b. Al GHE vûl durmît

(50) a. La çhosse la ul fate

b. La çhosse la GHE ul fate

5. Conclusions

In this article we have considered the different syntactic possibilities exhibited by deontic

want in the eastern Venetan variety of Fossalta di Piave. Some of these syntactic properties

are not attested in the neighbouring varieties but they have an interesting counterpart in the

syntactic behavior of deontic want in Friulian. From a closer comparison between Fossaltino

and Friulian it emerged that the deontic construction with the modal want and a past participle

is of two types: in one case, a passivisation process is involved, in the other, the past

participle conveys an aspectual value.

The two construction can be distinguished by the following: (i) the type of subject (ii)

the past participial agreement, (iii) the type of lexical verb admitted and (iv) the distributional

properties of the clitic ghe. More precisely, the passive construction exhibits a referential

clitic subject (also a DP subject can be present) and past participial agreement. In this case,

the construction is admitted only with transitive verbs and the clitic ghe may also have a

benefactive reading, as the complementary distribution with other benefactive clitic shows.

In the participial complement construction we have argued for the presence of a null

expletive subject (a lexicalized expletive is present in Friulian) and there is no participial

agreement. This construction is possible with all types of verbs and the clitic ghe has no

benefactive reading. Furthermore, Fossaltino deontic want can be followed by a that-clause.

We have proposed that also in this case a null expletive clitic is present. We tentatively

suggest that the presence of a null expletive could be responsible for the more impersonal

reading native speakers attribute to ghe vol plus participial complement.

References

Anderson, John M. (1980). „Anti-Unaccusative, or: Relational Grammar in Case Grammar‟

Revue Roumaine de Linguistique 25/3: 193-225.


61

ASIt = Atlante Sintattico d‟Italia, http://www.asis-cnr.unipd.it/

Benincà, Paola & Cecilia Poletto (1994). „Bisogna and its Companions‟ in G. Cinque et al.,

Paths Towards Universal Grammar. Studies in Honor of Richard S. Kayne.

Washington D.C.: Georgetown University Press, pp. 35-57.

Benincà, Paola & Cecilia Poletto (1997). „The diachronic development of a modal verb of

necessity‟ in A. van Kemenade & N. Vincent, Parameters of Morphosyntactic

Change. Cambridge: Cambridge University Press, pp. 94-118.

Benincà, Paola & Christina Tortora (2009). „Towards a Finer-Grained Theory of Italian

Participial Clausal Architecture‟ Proceedings of the 32nd

Annual Penn Linguistics

Colloquium,15/1: 17-26; http://repository.upenn.edu/pwpl

Berizzi, M. (2011). “Usi deontici di toccare nelle varietà italiane: considerazioni storiche”,

handout presented at CIDSM 6, University of Cambridge, 16-17 June 2011.

Cinque, Guglielmo (1999). Adverbs and Functional Heads. A Cross-Linguistic Perspective.

Oxford-New York: Oxford University Press.

Cinque, Guglielmo (2004). „ 'Restructuring' and Functional Structure‟ in A. Belletti,

Structures and Beyond. The Cartography of Syntactic Structures, vol. 3. New York:

Oxford University Press,.pp. 132-191.

Cruschina, Silvio (2010). „Aspetti morfologici e sintattici degli avverbi in siciliano‟ in J.

Garzonio, Quaderni di Lavoro ASIt, 11. Studi sui dialetti della Sicilia. Padova:

Unipress, pp. 21-42.

Freeze, Ray (1992). „Existentials and Other Locatives‟ Language 68/3: 553-595.

Kayne, Richard (2007). „Expletives, Datives, and the Tension between Morphology and

Syntax‟, MS, New York University;

http://linguistics.as.nyu.edu/object/RichardSKayne.html.

Ledgeway, Adam (2000). A Comparative Syntax of the Dialects of Southern Italy: A

Minimalist Approach. Oxford-Boston: Blackwell.

Nazzi, Gianni (2003). Vocabolario Italiano-Friulano, Friulano-Italiano. Udine: Clape

Culturâl Acuilee.

Parry, Mair (2010). “Non-canonical subjects in the early Italian vernaculars”, MS, University

of Bristol.

Patruno, Barbara (2005). Il verbo volere nelle varietà italo-romanze. Tra sintassi e semantica.

PhD Thesis. Università di Padova.

http://www.asis-cnr.unipd.it/

http://linguistics.as.nyu.edu/object/RichardSKayne.html

Berizzi - Rossi

62

Remberger, Eva-Maria (2006). „Syntax and Semantics of the deontic WANT-passive in Italo-

Romance‟ in B. Lyngfelt & T. Solstad, Demoting the Agent. Passive, middle and other

voice phenomena. Amsterdam-Philadelphia: John Benjamins, pp. 249-274.

Rohlfs, Gerhard (1969). Grammatica storica della lingua italiana e dei suoi dialetti. Sintassi e

formazione delle parole. vol. 3. Torino: Einaudi.

Tortora, Christina (1997). The Syntax and Semantics of the Weak Locative, PhD thesis,

University of Delaware.

Deontic ghe vol with past participle in some varieties of ...

Documents

Transcript of Deontic ghe vol with past participle in some varieties of ...